Attachment in Surinam-Dutch Families:
A Contribution to the Cross-Cultural Study
of Attachment
V Marinus H. van IJzendoorn
Center for Child and Family Studies, Leiden University, The Netherlands
A study on attachment in indigenous Dutch families and in families who came to the Netherlands from Surinam—a former Dutch colony in South America is described. Sixty-five mothers and their infants participated, including 26 mothers who had immigrated from Surinam. Attachment be-haviours and maternal responsiveness were recorded in the Strange Situation and in free play. The Infant Characteristics Questionnaire, and a question-naire on child-rearing attitudes were completed. In the Surinam-Dutch group, maternal responsiveness was related to quality of attachment, although the Surinam-Dutch mothers scored significantly lower on the responsiveness scale than the Dutch mothers. The attachment classification distribution of the Surinam-Dutch dyads did not significantly differ from Dutch or global distributions. However, Surinam-Dutch mothers appeared to be more anxious about child-rearing than Dutch mothers. Surinam-Dutch mothers who had recently arrived in Holland tended to show less responsive-ness, and were more anxious about child-rearing than Surinam-Dutch mothers who had immigrated several years ago. These data may reflect strains that arise from the transition from one culture to another.
INTRODUCTION
t In attachment theory it is hypothesised that the readiness to establish an attachment relationship, defined äs a relatively durable affective
relation-Requests for reprints should be sent to Marinus H. van IJzendoorn, Center for Child and Family Studies, Leiden University, P. O. Box 9555, NL-2300 RB Leiden, The Netherlands. The author gratefully acknowledges the assistance of Marion van Dam in scoring the responsiveness scale, and Hans Plomp for coding part of the Strange Situations. The help of Cyril Fräser, Regina van den Berg, Alice Gritter, Corinne van de Heuvel, Ria Pengel, and Linda den Uijl in collecting the data was indispensable. Preparation of this paper was supported by a senior Fulbright Fellowship at the National Institute of Child Health and Human Development (Bethesda, U.S.A.).
ship between children and one or more specific persons with whom they regularly interact (Bowlby, 1971; Ainsworth, Blehar, Waters, & Wall, 1978), is a universal phenomenon. Because of its evolutionary roots, attachment is claimed to be present in all human beings, irrespective of the specific culture in which they were born (Ainsworth, 1967; Bowlby, 1971). Studies in several different cultures have been carried out to test the universality hypothesis, not only in the U.S.A. but also in Western-European countries like Great Britain (Smith & Noble, 1987), The Nether-lands (Van IJzendoorn, Goossens, Kroonenberg, & Tavecchio, 1985), Sweden (Lamb, Hwang, Frodi, & Frodi, 1982), and West Germany (Grossmann, Grossmann, Huber, & Wartner, 1981); in Africa (Kermoian & Leiderman, 1986); in Israel (Sagi et al. 1985); and in Japan (Durrett, Otaki, & Richards, 1984; Miyake, Chen, & Campos, 1985). From these studies it can be deduced that attachment behaviours are present in every culture and that the intracultural differences in quality of attachment—äs measured through the Strange Situation procedure (Ainsworth & Wittig, 1969)—are larger than the cross-cultural differences (Van IJzendoorn & Kroonenberg, 1988).
Cross-cultural comparisons could not take into account, however, South-American cultures, because attachment research has not yet been carried out in that part of the world (Van IJzendoorn & Kroonenberg, 1988). Following Li-Repac (1982), who studied attachment in Chinese Immigrant families in the U.S.A., we decided to study attachment in families who came to the Netherlands from Surinam—a former Dutch colony. Until the beginning of the seventies, the Surinamese emigrated mainly because they wanted to get a higher level of education in the Netherlands. Most of these emigrants were Creole-Surinamese. In the years before and after the independence of Surinam in 1975 the number of emigrants grew enormously because of the negative economic prospects in Surinam, and because of the insecure political Situation there. Hindustani, Javanese, and other ethnic groups decided to emigrate to the Netherlands especially because they feared Creole domination after independence. In 1985 more than 30% of the whole Surinamese population was residing in the Netherlands (about 195 000 people) (Van den Berg-Eldering, 1988).
able to answer these questions, however, the validity of the Strange Situation procedure to measure attachment in Surinam-Dutch families must be established. Parental responsiveness to infant's signals is consi-dered to be one of the most important validity indices for attachment security (Ainsworth et al., 1978; Bretherton, 1985; Lamb, 1987). It was therefore hypothesised that in the Surinam-Dutch sample a higher degree of maternal responsiveness would correspond with greater infant attach-ment security.
In sum, four hypotheses will be tested. First, we hypothesise that the Strange Situation procedure is a valid Instrument to measure attachment security in our Surinam-Dutch sample, and correlates with maternal responsiveness. Secondly, we hypothesise that the Surinam-Dutch attach-ment classification distribution does not differ significantly from the distri-bution in a comparable Dutch sample nor from the "global" distridistri-bution (Van IJzendoorn & Kroonenberg, 1988). Thirdly, we suppose that intra-cultural differences in attachment classification distribution may be greater than cross-cultural differences. Fourthly, it is hypothesised that Surinam-Dutch mothers have not been completely assimilated into the Surinam-Dutch culture, but on the contrary show different child-rearing behaviours and attitudes from Dutch mothers.
METHOD Subjects
In this study 65 mothers and their infants participated. All subjects had Dutch nationality, but 26 mothers had immigrated from the former Dutch colony, Surinam, a South-American country. Mean duration of stay in Holland was 13.2 years. (SD = 4.89; minimum 6 years and maximum 22 years). Fifteen mothers from the Surinam-Dutch sample were Creole, eight were Hindu, and three were from other ethnic groups. Mean educa-tional level of the Surinam-Dutch mothers was 4.4 (SD = 1.63) on a scale ranging from l (low, i.e. 6 years of school attendance) to 9 (high, i.e. 16 years or more). Their infants were 18 months old (SD = 2.10) and 39% of the infants were boys. Mean educational level of the 39 Dutch mothers was 6.2 (SD = 2.62). The Dutch infants were 18.7 months old (SD = 0.61), and 49% were boys. Subjects were recruited through birth announcements in a local paper; and through personal contacts of research assistants.
Procedures
shortened, however, if the baby is very upset. The infant is confronted with a stränge environment (i.e.the playroom in the laboratory), with a Strange person, and with two separations from the mother (for details see Ains-worth et al., 1978). To assess the quality of the attachment relationship, the behaviour of the infant is scored on six 7-point rating scales. The scales are for proximity and contact seeking, maintenance of contact, resistance, avoidance, search behaviour, and distance interaction. The pattern of scores in the two reunion episodes leads to the classification of mother-infants dyads äs having anxiously avoidant attachment (A), secure attach-ment (B), or anxiously resistant attachattach-ment (C).
Two native Dutch observers (Hans Plomp and Marinus van IJzendoorn) independently coded 20 video-recorded Strange Situations. Mean reliabil-ity of the interactive scales was 0.78; intercoder agreement for classification in the three main categories was 90%. The author of this paper, who coded all Strange Situations, was trained by Brian Vaughn in a Strange Situation workshop at the University of Minnesota (Minneapolis). After coding crying behaviour in the two reunion episodes following the coding system of Ainsworth et al. (1978) (intercoder reliability 0.99), a continuous scale for security of attachment was computed, based on the algorithm validated by Richters, Waters, and Vaughn (1988).
Responsiveness. Responsiveness was defined äs the mother's ability to perceive and Interpret accurately signals implicit in the infant's behaviour, and to respond to them promptly and accurately (Ainsworth, Bell, & Stayton, 1974). Mothers were asked to play with their infant, äs they habitually did at home, in our laboratory playroom for 8 minutes before the Strange Situation procedure started. Play sessions were recorded on video, and Ainsworth et al's (1974) rating scale for measuring responsive-ness was applied to rate maternal behaviour. The native Dutch coder— Marion van Dam—was blind to infants' attachment classifications. Inter-coder reliability on ten sessions was 0.92.
Infant Characteristics Questionnaire (ICQ). The mothers in our sam-ples completed the Dutch Version of the ICQ (Bates, Freeland, & Louns-bury, 1979) to assess temperament. Kohnstamm (1984) found satisfactory validity and reliability figures in a sample of 7000 Dutch parents. The following dimensional structure was found (see also Tavecchio & Van IJzendoorn, 1987):
1. Difficultness. This dimension indicates how often a child cries or fusses and how much attention he/she generally demands from the care-giver. A low score indicates an "easy" temperament. Alpha reliability for the Surinam-Dutch mothers was 0.63, and for the Dutch mothers 0.69.
new circumstances, such äs strangers or environments. A low score indi-cates a high degree of adaptability. Alpha reliability for the Surinam-Dutch sample was 0.75 and for the Surinam-Dutch mothers 0.87.
3. Cuddly. This dimension reflects the child's acceptance of being picked up or cuddled. A low score indicates dislike for being cuddled. Alpha reliability for Surinam-Dutch mothers was too low: 0.12, and for Dutch mothers 0.66.
4. Persistence. This dimension indicates the child's obedience and docility. A high score corresponds with disobedience. Alpha reliability for the Surinam-Dutch sample was 0.73, and for the Dutch sample 0.76.
5. Mood. The child's typical mood (cheerful or rather sober). High scores point to a rather dejected mood. Alpha reliability for the Surinam-Dutch group was 0.74, and for the Surinam-Dutch group 0.86.
Child-rearing attitudes. A questionnaire measuring maternal attitude to child-rearing methods and problems was applied. The questionnaire has been developed and validated by Engfer and Schneewind (1976) who found the following scales:
1. Stress. This scale, consisting of six items, indicates the degree to which the mother feels overburdened by child-rearing problems. Alpha reliability for the Surinam-Dutch group was 0.77; and for the Dutch mothers 0.67.
2. Punishment. This scale of six items reflects mother's opinion about physical punishment äs a child-rearing method. Alpha reliability for the Surinam-Dutch sample was 0.74 and for the Dutch mothers 0.70.
3. Frustration. This scale of five items indicates parental feelings of being restricted by the child in doing things for themselves. Alpha reliabil-ity in the Surinam-Dutch group was 0.69, and for the Dutch mother 0.78. 4. Anxiety. This scale is based on eight items and reflects maternal concern for the child's health and development. Alpha reliability for the Surinam-Dutch mothers was 0.64 and for the Dutch group 0.72.
5. Role. This scale, consisting of four items, points to the role of children in the mother's life: the child's importance for maternal feelings of self-fulnlment. Alpha reliability for the Surinam-Dutch group was 0.65, and for the Dutch mothers 0.82.
Analysis
First, a Spearman correlation coefficient for the relation between attach-ment security and maternal responsiveness in the Surinam-Dutch sample was computed. Because of the small sample size a distribution-free coef-ficient was preferred. Secondly, intracultural differences in attachment security, responsiveness, temperamental variables, and child-rearing atti-tudes within the Surinam-Dutch sample were tested through the distribution-free Mann-Whitney U test. Thirdly, cross-cultural differences in attachment classification distributions were assessed using a test for multinomial distributions, in which proportions found for the global and the Dutch samples are taken äs the expected values for the Surinam-Dutch sample. Furthermore, using Van Uzendoorn and Kroonenberg's (1988) algorithm, the Surinam-Dutch sample was projected into the global distri-bution to see which samples were closest to the Surinam-Dutch sample. Lastly, temperamental differences and differences in child-rearing attitu-des and behaviour between the Surinam-Dutch and the Dutch dyads were tested through analyses of covariance.
RESULTS AND DISCUSSION
The Strange Situation and Surinam-Dutch families. One of the most important indicators of the validity of the Strange Situation procedure is its correspondence with maternal responsiveness. In this study, responsive-ness of Surinam mothers correlated 0.34 (P < 0.05) with the continuous variable security äs computed according to the Richters et al. (1988) algorithm. The more responsive Surinam mothers were to their infants' signals in a free-play Situation, the more secure their infants behaved in the stressful Strange Situation procedure. Although the correlation is not high compared to the Ainsworth et al. (1978) findings, it is well above the average correlational effect size found by Goldsmith and Alansky (1987) in their meta-analysis of 12 pertinent studies (0.16). Responsiveness tended to correlate positively with time spent in Holland (0.35; P = 0.059; N = 21): the longer the Surinam-Dutch mothers had stayed in Holland, the more responsive to their infants' signals they tended to be.
TABLE 1
Comparison of Surinam-Dutch, Dutch, and Global Attachment Classifica-tion DistribuClassifica-tions
Surinam—
Dutch Dutch Global1
Classification Avoidant (A) Secure (B) Resistant (C) Total 6 14 6 26 (23) (54) (23) (100) 14 21 4 39 (36) (54) (10) (100) 423 1294 273 1990 (21) (65) (14) (100)
'Global distribution derived from Van Uzendoorn and Kroonenberg (1988).
Cross-cultural differences in attachment classification distributions are described in Table l. The Surinam-Dutch sample seems to have contained an overrepresentation of resistant infants (23% versus 14% in the global distribution) but a statistical comparison of the Surinam-Dutch distribu-tion with the Dutch and the global distribudistribu-tions did not reveal any significant differences (Surinam-Dutch versus global distribution: X2 (2) =
2.14; P =0.34; Surinam-Dutch versus Dutch distribution: X2 (2) = 2.38; P
= 0.30).
Using the following formulae for Computing the X ordinate and Υ co-ordinate respectively, the Surinam-Dutch sample can be projected into the correspondence analysis solution of the samples-by-classification table given in Van Uzendoorn and Kroonenberg (1988):
χ = (-0.6447 fA - 0.0313 fB + 1.1473 fc)/(0.2696 X N) γ = (-0.7031 fA + 0.3495 fB - 0.5670 fc)/(0.2286 X N) where N = fA + fB + fc.
The X co-ordinate is 0.3677, and the Υ co-ordinate is -0.4589. The Surinam-Dutch sample distribution was closest to the U.S. samples of Schneider-Rosen and Cicchetti (1984), and Egeland and Farber (1984), and the Chinese-American Li-Repac (1982) sample (see Van Uzendoorn & Kroonenberg, 1988). These samples were characterised by an overrepre-sentation of anxious attachments, especially C-type attachments.
Differences between Surinam-Dutch and Dutch dyads. Although the
TABLE 2
Temperamental and Attitudinal Differences Between Surinam-Dutch and Dutch Mother-lnfant Dyads Variables Surinam-Dutch (n = 26) Dutch (n = 39) M (SD) M (SD)
Note. Degrees of freedom for the F tests: 1,63.
Controlled for educational level F P Temperament Difficulties Adaptability Cuddly Persistence Mood Attitudes Stress Punishment Frustration Anxiety Role Responsiveness Security 3.4 3.1 5.0 4.8 2.0 2.1 1.9 2.3 2.2 1.5 3.6 0.00 (1.03) (1.32) (1.57) (1.74) (1.04) (0.65) (0.77) (0.74) (0.49) (0.59) (2.14) (0.82) 3.0 2.8 4.6 4.3 2.2 2.4 2.4 2.5 2.7 2.1 6.5 0.00 (0.67) (1.06) (1.20) (1.23) (0.85) (0.50) (0.49) (0.51) (0.47) (0.64) (1.70) (1.20) 4.3 3.2 1.3 3.1 0.3 5.3 6.2 5.3 12.4 6.6 25.6 0.01 0.043 0.077 0.264 0.082 0.580 0.024 0.015 0.025 0.001 0.012 0.000 0.931
Standard deviations for maternal responsiveness in the Surinam-Dutch and Dutch samples are given. An analysis of covariance with maternal educational level äs covariate (Surinam-Dutch and Dutch mothers dif-fered significantly on educational level, ? (62) = 3.16; P < 0.01), showed that Surinam-Dutch mothers displayed significantly less responsiveness compared to Dutch mothers.
(Pillai = 0.12; F (4,58) = 2.02; P = 0.10). Time spent in Holland only correlated significantly with anxiety (0.60; P = 0.002; N = 21). Surinam-Dutch mothers appeared to be more anxious about child-rearing if they had spent less time in Holland.
It should be kept in mind that the sizes of the Surinam-Dutch and Dutch samples were relatively small, and that these samples cannot be considered representative of the two populations. Generalisations to all Surinam-Dutch and Surinam-Dutch families are, therefore, not warranted until our results have been replicated in larger groups.
CONCLUSIONS
Because maternal responsiveness and infant's attachment security corre-late in the expected direction, the Strange Situation appears to tap validly an important dimension of the relationship between Surinam-Dutch mothers and their 18-month-old infants. Furthermore, the Strange Situa-tion classificaSitua-tion distribuSitua-tion does not show signiflcant intracultural or cross-cultural differences. The procedure appears to be rather robust for cultural variations in child-rearing practices and attitudes (Van Uzen-doorn, 1990). Surinam-Dutch and Dutch mothers do differ significantly in child-rearing behaviour, i.e. responsiveness, and in their attitudes to the bürden of rearing a young child. Surinam-Dutch mothers show less responsiveness to infants' signals, or they experience child-rearing äs a heavier bürden compared with Dutch mothers. They favour physical punishment more, and they perceive their infants temperamentally äs more difficult than Dutch mothers. Although these cultural differences exist, even when differences in educational level have been controlled for, the same percentage of Surinam-Dutch and Dutch infants was securely attached to their mothers (54%).
ambivalent-resistantly attached children (Amsworth et al., 1978), in Li-Repac's sample äs well äs in our Surinam-Dutch sample. We found that Surinam-Dutch mothers who had lived in Holland for a considerable number of years tended to show more responsiveness, and were definitely less anxious about their infants' development than Surinam-Dutch mothers who had immigrated more recently. Our data therefore appear to indicate that strains may arise from a transition from one culture to another. Neverthe-less, in both Li-Repac's and our sample most infants appeared to adapt rather well to the cultural transition and to be robust for their mothers' suboptimal responsiveness, indicating that infants' attachment behaviour should be considered äs a conditional strategy flexibly adaptable to the specific sociocultural niche in which children have to live (Hinde, 1982; Lamb, Thompson, Gardner, & Charnov, 1985).
Manuscnpt accepted 18 December 1989
REFERENCES
Ainsworth, M D S (1967) Infancy m Uganda Infant care and the growth of love Baltimore, Md Johns Hopkins Umversity Press
Ainsworth, M D S , Bell, S M , & Stayton, D J (1974) Infant-mother attachment and social development 'socialization' äs a product of reciprocal responsiveness to Signals In H P M Richards (Ed ), The Integration of a chüd mto a social world London Cambridge Umversity Press
Ainsworth, M D S , Blehar, M C , Waters, E , & Wall, S (1978) Patterns of attachment, a
psychologtcal study ofthe Strange Situation Hillsdale, NJ Lawrence Erlbaum Associates
Ine
Amsworth, M D S & Wittig, B A (1969) Attachment and exploratory behavior of one-year-olds m a Strange Situation In B M Foss (Ed ), Determmants ofmfant behavior, IV London Methuen
Bates, J E , Freeland, C A , & Lounsbury, M L (1979) Measurement of mfant difficult-ness Chüd Development, 50, 794-803
Bowlby, J (1971) Attachment and loss Vol I Attachment Harmondsworth Pengum Bretherton, I (1985) Attachment theory Retrospect and prospect In I Bretherton & E
Waters (Eds ), Growmg pomts of attachment theory and research Monographs of the
Society for Research m Child Development, 50, (1-2, Senal No 209), 3-37
Dürre«, M E , Otaki, M , & Richards, P (1984) Attachment and the mother's perception of support from the father International Journal of Behavioral Development, 7, 167-176
Egeland, B & Farber, E A (1984) Infant-mother attachment Factors related to its development and changes over time Chüd Development, 55, 753-771
Goldsmith, H H & Alansky, J A (1987) Maternal and mfant temperamental predictors of attachment A meta-analytic review Journal of Consulting and Chmcal Psychology, 55, 805-816
Hinde, R A (1982) Attachment Some conceptual and biological issues In J Stevenson-Hmde & C Murray Parkes (Eds), The place of attachment m human behaviour New York Basic Books (Pp 60-78 )
Kermoian, R & Leiderman, P H (1986) Infant attachment to mother and child caretaker m an East Afncan Community International Journal of Behavioral Development, 9, 455— 469
Lamb, M E (1987) Predictive imphcations of individual differences in attachment Journal
of Consulting and Chnical Psychology, 55, 817-824
Lamb, M E , Hwang, C P , Frodi, A ,& Frodi, M (1982) Secunty of mother- andfather-infant attachment and its relation to sociabihty with strangers in traditional and non-traditional Swedish families Infant Behavwr and Development, 5, 355-367
Lamb, M E , Thompson, R A , Gardner, W , & Charnov, E L (1985) Infant-mother
attachment The origins and developmental sigmficance of individual differences m Strange Situation behavior Hillsdale, NJ Lawrence Erlbaum Associates Ine
Li-Repac, D C (1982) The impact of acculturation of the child-rearmg attitudes andpractices
of Chinese-American families Consequences for the attachment process (Doctoral
disser-tation, Umversity of California Berkeley)
Miyake, K , Chen, S J , & Campos, J J (1985) Infant temperament, mother's mode of interaction, and attachment m Japan An mtenm report In I Bretherton & E Waters (Eds), Growing pomts of attachment theory and research Monographs ofthe Society for
Research m Child Development, 50, (1-2, Senal No 209), 276-297
Richters,! E , Waters,E , & Vaughn, B E (1988) Empincalclassificationof infant-mother relationships from mteractive behavior and crying dunng reumon Child Development, 59, 512-522
Sagi, A , Lamb, M E , Lewkowicz, K S , Shoham, R , Dvir, R , & Estes, D (1985) Secunty of infant-mother, -father, metapelet attachments among Kibbutz-reared Israeli children In I Bretherton & E Waters (Eds), Growing pomts of attachment theory and research Monographs ofthe Society for Research in Child Development, 50 (1-2, Senal No 209), 257-275
Schneider-Rosen, K & Cicchetti, D (1984) The relationship between affect and cognition in maltreated infants Quality of attachment and the development of Visual self-recognition
Child Development, 55, 648-658
Smith, P K & Noble, R (1987) Factors affecting the development of caregiver-mfant relationships In L W C Tavecchio & M H van IJzendoorn (Eds), Attachment insocial
networks Contributions to the Bowlby-Ainsworth attachment theory Amsterdam
Else-vier Science Pubhshers (Pp 93-134 )
Tavecchio, L W C & Van IJzendoorn, M H (1987) Perceived secunty and extension of the child's rearmg context A parent-report approach In L W C Tavecchio & M H van IJzendoorn (Eds), Attachment m social networks Amsterdam Eisevier Science Pub-hshers (Pp 35-92 )
Van IJzendoorn, M H (1990) Developments m cross-cultural research on attachment A methodological note Human Development, 33, 3-10
Van IJzendoorn, M H , Goossens, F A , Kroonenberg, P M , & Tavecchio, L W C (1985) Dependent attachment B-4 children m the Strange Situation Psychological
Re-ports, 57, 439-451
REFERENCE NOTES
1 Engfer, A & Schneewind, K A (1976) Ein Fragebogen zur Erfassung selbstperzipierter
elterlicher Erziehungsemstellungen [A questionnaire for the assessment of self-perceived
parental attitudes towards child-rearmg ] Unpubhshed manuscnpt (Arbeitsbericht 9, EKB-projekt), Umversity of Trier, FRG
2 Kohnstamm, G A (1984, April) Bates' Infant Characteristtc Questionnaire (ICQ) m the
Netherlands Paper presented at the fourth Biennial International Conference on Infant
Studies, New York
